Disclosures. Objectives. CBCT in Dentistry and Research 4/17/2017. Protocols for CBCT Imaging

Transcription

1 Assessing Alveolar Bone Height and Thickness Using Cone Beam Computed Tomography: Are Looks Deceiving? American Association of Orthodontists April 25, 2017 David A. Covell, Jr., DDS, PhD Professor and Chair Department of Orthodontics Oregon Health & Science University Portland, Oregon Objectives Investigate CBCT accuracy and precision for assessing alveolar bone Part 1: Cadavers- alveolar bone height and thickness Compare CBCT- derived measurements to direct measurements Compare accuracy of measurements using varying CBCT settings Part 2: Patients undergoing periodontal surgery procedures Compare measurements of alveolar bone height- CBCT vs. direct Evaluate ability to identify root dehiscences and fenestrations Conclusions & implications on use of CBCT in assessing presence/absence of alveolar bone Disclosures No association with any products Note: various brands of CBCT units will be mentioned Some older generation, others newer No intent to suggest one brand is superior to another Reporting data from the literature or CBCT units available at the Oregon Health & Science University School of Dentistry i-cat Carestream CS 9300 CBCT in Dentistry and Research 2D and 3D volume renderings * Sub-millimeter spatial resolution ** Optimal acquisition settings? Orthodontics: alveolar bone morphology Pre-treatment Changes with treatment Long-term High or low resolution scans? Balance image detail & radiation exposure * Ballrick et al, Am J Orthod Dentofac Orthop 2008; 134: ** Pauwels et al., Eur J Radiol 2011; 81: article: Protocols for CBCT Imaging Multiple acquisition settings Impact image quality, effective radiation dose Duration of scan and number of projection images * Size of field of view ** [4-17 cm] Voxel size ** [ mm] Image quality vs. effective radiation dose ALARA As low as reasonably achievable (1999)*** ALADA As low as diagnostically acceptable (2014)**** National Council on Radiation Protection and Measurements * Brown et al., Angle Orthod 2009; 79:150-7 ** Ludlow et al., Dentomaxillofac Radiol 2006;35: *** Martin et al., Appl Radiat Isot 1999;50:1-19 **** NCRP Settings for i-cat Imaging Sciences International, Hatfield, PA 1

2 Part 1: Facial Bone Height and Thickness Measurements in Cadaver Study (Adam Timock, Lane Cook) Comparison of alveolar bone measurements: Made from CBCT scans acquired using i-cat CBCT unit s default setting, and settings aimed to produce lesser or greater resolution Compare CBCT to direct measurements via dissection Sample Selection Protocol reviewed by OHSU s IRB 17 cadavers accessed through Department of Integrative Biosciences Screening inclusion criteria: Teeth with periodontium free of damage Sample teeth and adjacent teeth having no metal restorations Sample selected: 12 cadavers: 5 female, 7 male; all Caucasians Mean age: 77 years (range: years) 65 teeth selected: 48 anterior, 17 posterior teeth Timock et al., Am J Orthod Dentofac Orthop 2011;140: Cook et al. Orthod Craniofac Res 2015;18(Suppl.1): Timock et al., Am J Orthod Dentofac Orthop 2011;140: Cook et al. Orthod Craniofac Res 2015;18(Suppl.1): CBCT Settings 3 scans: default, 2 custom settings Long scan (LS)**: 26.9 seconds; 619 projection images; 360 o rotation; 0.2 mm voxel size Default scan (DS)*: 8.9 seconds; 309 projection images; 360 o rotation; 0.3 mm voxel size Short scan (SS)**: 4.8 seconds; 169 projection images; 180 o rotation; 0.3 mm voxel size Compared to direct measurements from dissection i-cat CBCT Unit Imaging Sciences International Direct Measurement of Bone Height Following acquisition of CBCT scans: Gingiva dissected Buccal bone height measurement (BBH) Digital caliper (0.01 mm) Distance from coronal-most point of crown to buccal alveolar crest, following the long axis of tooth Timock et al., Am J Orthod Dentofac Orthop 2011;140: Cook et al. Orthod Craniofac Res 2015;18(Suppl.1): Timock et al., Am J Orthod Dentofac Orthop 2011;140: Direct Measurement of Bone Thickness Buccal bone thickness measurement (BBT) Notch dissected in alveolar crest - 3 mm height Measured with a customized depth gauge (0.01 mm) Linear distance from cementum to lateral bone surface Document site of BBT measurement With caliper, measure from cusp tip to apical base of dissection site Defined the location where BBT measurements made in CBCT images Thickness measurements Made by 2 calibrated investigators (AT, LC) Repeated 3 times, at least one day apart CBCT Measurements DICOM files imported into Dolphin 3D Imaging Measurements: 0.5 mm slice thickness oriented using a standardized protocol BBH: measured in sagittal plane, parallel to long axis of tooth BBT: measured in axial plane at vertical location where direct BBT measurements had been made with dissections Measurements made independently by 2 calibrated investigators (AT, LC) Repeated 3 times, at least 1 day in between 2

3 Difference [Direct CBCT] (mm) Difference [Direct CBCT] (mm) Difference [Direct CBCT] (mm) Difference [Direct CBCT] (mm) Difference [Direct CBCT] (mm) 4/17/2017 Statistical Analysis Comparison of measurements 3 repeat measurements averaged Direct vs. CBCT mean values compared using paired t-tests (p<.05) ANOVA: Comparison of measurements from 3 CBCT settings Concordance correlation coefficients (CCC): Rater & measurement comparisons Bland-Altman plots * : Agreement of CBCT vs. direct measurements Results: Direct Measurements Both intra- and inter-rater reliability were very high * Bland & Altman. Stat Methods Med Res 1999;8: Bland-Altman Plot: Example Bone Height, Long Scan - CBCT scan - Dissection - 65 teeth measured Bland-Altman Plots: Bone Height- Long Scan & Short Scan o Each circle = 1 tooth (n=65) Y-axis: difference between direct and CBCT measurements X-axis: average of the direct and CBCT measurements (X-axis) Average [Direct & CBCT] (mm) - - Dashed line = mean difference among all direct and CBCT measurements Solid lines = 95% limits of agreement Long Scan Short Scan LS: mean difference = mm SS: mean difference = 0.00 mm 95% LOA to 0.40 mm 95% LOA to 1.04 mm Cook et al. Orthod Craniofac Res 2015;18(Suppl.1): Bland-Altman Plots: Bone Thickness- Long Scan & Short Scan Measurement Comparisons Long Scan Short Scan LS: mean difference = 0.00 mm SS: mean difference = mm 95% LOA to 0.24 mm 95% LOA to 0.28 mm Cook et al. Orthod Craniofac Res 2015;18(Suppl.1): Accuracy and precision: Direct measurements: high reliability Correlation coefficients close to 1.0 Valid for use as controls CBCT-derived measurements: accurate, but varying precision Accuracy: accurate with all 3 settings (no difference statistically from direct measurements) Precision: greater with longer scans as shown by Bland-Altman 95% LOA: E.g., bone height: Long Scan: mm Default Scan: mm Short Scan: mm Timock et al., Am J Orthod Dentofac Orthop 2011;140: Cook et al. Orthod Craniofac Res 2015;18(Suppl.1): LS SS 3

4 Long vs. Short Scan BBH and BBT: Long, Short & Default Scans Variable BBH (mm) BBT (mm) BBH CCC BBT CCC Long Scan Default Scan t Short Scan Accuracy: both short & long scans shown to be accurate Precision: short scan images had less precision than long scan Correlations with direct measurements: BBH - strong: CCC: > 0.97 BBT - moderate to poor: CCC: > 0.86 Overall: agreement higher for bone height than for bone thickness Timock et al., Am J Orthod Dentofac Orthop 2011;140: Cook et al. Orthod Craniofac Res 2015;18(Suppl.1): Previous Studies Using Skulls or Cadavers Cephalometric landmarks Brown et al. Linear accuracy of cone beam CT derived 3D images. Angle Orthod 2009;79:150-7 [U Louisville] Dried skulls (n=19), assessed accuracy of landmark identification Varied settings for number of projection images: 153, 306, 612 Findings: No statistical difference in measurement accuracy among the 3 settings Mean differences: 0.44, 0.38, and 0.32 mm, respectively Suggest can achieve 75% reduction in radiation dose using 153 relative to 612 projection images Radiation Dose Considerations Current study: Short Scan (180 o, 169 projection images) vs. Long Scan (360 o, 619 projection images) Likely reduce effective radiation dose by 73% Little compromise in measurement accuracy Comparison of CBCT units and settings: Pauwels et al. Effective dose range for dental cone beam computed tomography scanners. Eur J Radiol 2011;81: CBCT units, varied settings for field of view, tube output, other exposure factors Default settings: usv By changing settings: usv Changes in FOV had greatest impact on effective radiation dose Correlations: BBH & BBT Concordance correlation coefficients: favor BBH measurement Long Scan: BBH = 0.99; BBT = 0.94 Default settings: BBH = 0.98; BBT = 0.86 Short Scan: BBH = 0.97; BBT = 0.88 Why better results for measurements of bone height vs. thickness? Voxel size, landmark identification High contrast differential is best BBH: Enamel vs. air BBT: Cementum vs. bone Limitations of Skull/Cadaver Studies Older age subjects Stable patient positioning during scan Non-living tissue Impact of dehydration and embalming? Preferable to compare with CBCT assessments of alveolar bone height in living humans! Leung et al., Am J Orthod Dentofac Orthop 2010;137:S Hitachi, CB Mercuray; Serial # Case Western Reserve, September

5 Part 2: Alveolar Bone Height in Patients Scheduled for Periodontal Surgery (Megan Miller, Vanessa Browne Peterson: icat Andrew Peterson: Carestream CS 9300) Objectives: 1. Investigate accuracy and reliability of CBCT compared to direct measurements of facial alveolar bone height in vivo 2. Measurement of bone height only 3. Assess ability to detect naturally occurring dehiscences and fenestrations Methods and Patient Selection IRB approval Sample: Patients scheduled for surgery involving mucoperiosteal flaps Clinics: Periodontology; Oral and Maxillofacial Surgery CBCT imaging as part of patient s treatment plan CBCT images acquired within the previous 18 months Pilot study conducted to develop methodology Provided data for power calculation Surgical Procedures Osteotomy Implant Placement Sinus Lift Bone Graft Crown Lengthening Connective Tissue Graft Surgical Extraction Screening Sample Selection (Miller, Browne: i-cat 17-19) Inclusion Criteria CBCT scan on record in preparation for surgical procedure At least 18 years of age with permanent dentition Teeth in the area of surgery having natural tooth structure or small composite restorations Region of surgery free from osseous or soft tissue pathology Exclusion Criteria Alloy restorations (or implants) adjacent to or in the tooth of interest Previous surgery in the region 177 patients with CBCT images acquired September 2012 to July patients excluded 26 patients unavailable for Sx -Duplicate CBCT on same patient -Patient undergoing orthodontic treatment -Private practice referral -Patient did not undergo surgical procedure involving full thickness flap -Patient did not have intact tooth structure in surgical area 16 Patients (Mean age: 64 years) 32 teeth included Tooth Type Maxilla Mandible Total Anterior Central Incisor 6 6 Lateral Incisor 4 4 Canine Anterior total Posterior First Premolar Second Premolar Posterior total Total CBCT Acquisition and Measurements (Miller, Browne) CBCT scan: i-cat mm voxel; either: 10 x 10 cm (two arches), or 10 cm x 5 cm (one arch) DICOM File imported into Dolphin 3D Imaging for analysis Standardized protocol: 0.5 mm slice oriented to the M-D center of tooth and parallel to long axis 2 calibrated, blinded investigators 1. Measured bone height 2. Assessed presence or absence of dehiscence or fenestration Measurements/assessments made 3x Minimum 3 days between recordings 5

6 Difference: Direct-CBCT (mm) 4/17/2017 Dehiscence and Fenestration Dehiscence: alveolar bone height >3.0 mm from the CEJ Fenestration: bone defect identified toward apical end of root but not involving the alveolar crest Clinical Recordings Mucoperiosteal flap reflected 2-point bow compass recorded bone height Two calibrated investigators independently measured bone height Recorded presence or absence of a dehiscence or fenestration Photograph made of the surgical region JuniorDentist.com Results - Bone Height Measurements, Intra- and Inter-Rater Agreement Direct and CBCT measurements: Intrarater and interrater agreement was high CCC >0.98 As a result, measurements from the two raters were pooled for comparison of direct and CBCT measurements Results: CBCT vs. Direct Bone Height Measurements Relative to direct measurements, CBCT measurements showed larger values (reduced bone height) Difference: 2.3 mm; p<0.001 CBCT vs. direct measurement correlation coefficients showed poor agreement CCC: 0.64 Direct Measurement CBCT Measurement Mean Diff ± SD (mm) -0.09± ±0.99 Mean Abs ± SD (mm) 0.46± ±0.77 CCC (95% CI) 0.98 (0.96, 0.99) 0.98 (0.95, 0.99) Variable Direct CBCT Difference (direct - CBCT) Mean ± SD Mean ± SD Mean Mean Abs ± SD (mm) (mm) Diff ± SD (mm) (mm) FBH ± ± ± ± 2.61 CCC (95% CI) 0.64 (0.45, 0.77) Bland Altman Plot Measurement Differences by Tooth Type (International Tooth Numbering) Direct vs. CBCT measurements: Mean difference (CBCT-Direct): mm 95% level of agreement: to 3.20 mm Compare to cadaver short scan: Mean difference (CBCT-Direct): 0.0 mm 96% level of agreement: to 1.03 mm 4.0 Mean Diff = % LOA= (-7.76, 3.20) Average of Direct and CBCT (mm) Absolute difference, clinical vs. CBCT measurements: mm: 9 teeth Mx and Mn canines mm: 6 teeth Premolars (4); Mn incisors (2) mm: 5 teeth Premolars (4); lower incisor (1) mm: 12 teeth Mn incisors (7); Mn canines (3); Mn 1 st Premolars (2) Periodontal surgery patients 6

7 Dehiscences & Fenestrations 0.3 mm Voxel Arch Tooth # Clinical CBCT #1 CBCT #2 Mx 14 D D D Mx 13 None None F Mx 13 None None None Mx 13 None None None Mx 23 None None None Mx 23 None None None Mx 23 None None None Mx 24 D D D Arch Tooth # Dehiscences & Fenestrations Clinical CBCT #1 CBCT #2 Md L 34 None None F Md L 33 None None F Md L 33 D D D Md L 32 D D D Md L 32 F F F Md L 32 D D D Md L 32 None None None Md L 32 D D D Md L 32 F D D Md L 32 D None F Md L 31 D D D Arch Tooth # Clinical CBCT #1 CBCT #2 Md R 45 None None None Md R 44 None None D Md R 44 D D None Md R 44 None None None Md R 43 None None None Md R 43 D D D Md R 42 D D D Md R 42 D D D Md R 42 None D D Md R 41 D D D Md R 41 None D D Md R 41 D D D Dehiscences & Fenestrations Good Agreement Clinically: Agreement between raters: 100% 58% of teeth were found to have an alveolar bone defect 89% of defects: dehiscence 11% of defects: fenestration CBCT: Agreement between raters: 81% of the time Confirmed clinical findings: 72% of the time Defects either not detected or misidentified in CBCT Presence or absence of bone defect was misidentified in total of 9 teeth (28%), 8/32 (25%) were associated with mandibular teeth 6 Mn incisors; 2 Mn premolars 1 maxillary canine misidentified #35 Direct: 9.8 mm CBCT: mean 10.0 mm Both raters in agreement: No dehiscence/fenestration Dehiscence/Fenestration Fenestration #44 Direct: 10.9 mm; no dehiscence/fenestration CBCT rater 1: 10.5 mm; no dehiscence/fenestration CBCT rater 2: 15.6 mm; with dehiscence #43 Direct: 10.5 mm with fenestration CBCT: mean 18.2 mm with dehiscence 7

8 Periodontal Surgery CBCT Study Strengths Sample: Prospective study: standardized protocol Live patients Adequately powered CBCT: Cadaver vs. patients with icat 17-19: Same unit used for entire sample Raters calibrated Surgery: Able to directly visualize alveolar bone Raters calibrated Limitations Sample: Patients of older age (mean in 60 s) Periodontally compromised dentitions Time delay between CBCT and surgery (mean: 4.5 months) Only 2 teeth with fenestrations CBCT: 0.3 mm voxel: older CBCT unit (i- CAT 17-19) Limited experience of raters Surgery: Flap surgery can potentially damage bone Methods and Patient Selection (Peterson: Carestream CS 9300) IRB approval Sample inclusion: Age: 18 years or older Patients scheduled for surgery involving mucoperiosteal flaps Periodontology clinic CBCT images planned or on file within previous 18 months Teeth in area of interest have intact crowns Exclusion: Medically compromised Pregnancy Prisoner status Unable to consent Surgical Procedures Osteotomy Implant Placement Sinus Lift Bone Graft Soft Tissue Graft Crown Lenghtening Sample Selection (Peterson; CS 9300) CBCT Settings 26 patients screened from January to September patient excluded -CBCT Scan not acquire at 0.09 or 0.18 mm voxel Two calibrated raters assess CBCT & clinical measurements 25 Patients Mean age: 63 years 56 teeth included Carestream CS 9300 Localized area 5 x 5 cm field of view 0.09 mm voxel size 80 kvp/5 ma 360 o, 20 seconds scan time 524 mgy/cm 2 radiation dose Full dental arch 10 x 10 cm field of view 0.18 mm voxel size 90 kvp/4 ma 360 o, 8 seconds scan time 685 mgy/cm 2 radiation dose Carestream CS 9300 CBCT Carestream Health, Rochester, NY CBCT Acquisition and Measurements Measurement Comparisons DICOM File imported into Carestream s imaging software Standardized protocol: 0.18 mm slice oriented to the M-D center of tooth and parallel to long axis* 2 calibrated, blinded investigators 1. Measured bone height 2. Assessed presence or absence of dehiscence or fenestration Measurements/assessments made 3x Minimum 3 days between recordings Minimum 30 days between CBCT and surgery recordings Direct and CBCT Facial Bone Height Measurements for 0.09 Voxel Size (n=23) -Interrater agreement was high for direct measurements; moderate for CBCT Direct Measurement CBCT Measurement Mean Diff±SD (mm) 0.07± ±0.72 Mean Abs±SD (mm) 0.40± ±0.61 CCC (95% CI) 0.97 (0.92, 0.99) 0.94 (0.86, 0.97) Direct and CBCT facial bone height measurements for 0.18 voxel size(n=33) - Interrater agreement was high for both direct and CBCT measurements Direct Measurement CBCT Measurement Mean Diff±SD (mm) 0.0± ±0.45 Mean Abs±SD (mm) 0.42± ±0.32 CCC (95% CI) 0.98 (0.95,0.99) 0.98 (0.96,0.99) * Timock et al., Am J Orthod Dentofac Orthop 2011;140:

9 Difference: Direct-CBCT (mm) Difference: Direct-CBCT (mm) Difference: Direct-CBCT (mm) 4/17/2017 Bland-Altman Plots: CBCT vs. Direct Measurements Comparison: i-cat 17-19, 0.3 mm voxel Carestream CS mm voxel size Mean Diff = % LOA= (-1.88, 1.02) Average of Direct and CBCT (mm) Average of Direct and CBCT (mm) -6.0 Mean Diff = % LOA= (-7.76, 3.20) Voxel size= 0.09 mm Voxel size= 0.18 mm -8.0 Mean Diff = Mean Diff = % LOA= (-1.98, 0.86) 95% LOA= (-1.88, 1.02) Comparison: 0.18 mm voxel clinical 0.3 mm voxel short scan cadaver Mean Diff = % LOA= (-1.02, 1.04) Dehiscences and Fenestrations 0.09 and 0.18 mm voxel 0.09 mm voxel scans 23 teeth: Clinically: no teeth had fenestration or dehiscence CBCT: 6 teeth (26%) incorrectly identified as having either fenestration (2) or dehiscence (4) by one rater Mean Diff = % LOA= (-1.88, 1.02) Average of Direct and CBCT (mm) Dehiscences and Fenestrations 0.09 and 0.18 mm voxel 0.09 mm voxel scans 23 teeth: Clinically: no teeth had fenestration or dehiscence CBCT: 6 teeth (26%) incorrectly identified as having either fenestration (2) or dehiscence (4) by a rater 0.18 mm voxel scans 33 teeth: Clinically: 30 had no fenestration or dehiscence 2 teeth had fenestrations: With CBCT, 1 tooth fenestration correctly identified by both raters 1 tooth fenestration correctly identified by one rater 1 tooth had a dehiscence: With CBCT, correctly identified by both raters CBCT: 16 teeth (48%) incorrectly identified as having either fenestration (11) or dehiscence (5) by one or both raters Voxel Size Considerations 0.3mm voxel size 0.3mm resolution due to: Reduction in resolution due to noise, artifacts, and partial volume averaging Partial volume averaging effect: artifact generated when an object is imaged partially within 2 voxels CBCT image reflects the average density within each voxel, rather than the true value of the object Decreasing the CBCT voxel size has been suggested to improve the accuracy of linear measurements of alveolar Leung et al (2010): alveolar bone <0.6mm thick was invisible in CBCT, no matter the voxel size CB MercuRay, Hitachi Medical Systems, Twinsburg, OH Leung et al. Am J Orthod Dentofac Orthop 2010;137:S

10 Dihiscences & Fenestrations in Cadavers Improve accuracy with smaller voxel size? Patcas et al. Accuracy of cone-beam computed tomography at different resolutions assessed on the bony covering of mandibular anterior teeth. Am J Orthod Dentofac Orthop 2012;141:41-50 Cadaver study: CBCT* 0. 4 and mm voxel, followed by dissection Conclusion: Even with mm voxel scan, thin buccal bone was not reliably imaged Dehiscences & Fenestrations in Patients Liangyan et al: Class III patients (n=14 patients, 122 teeth) Accelerated osteogenic orthodontic surgery on maxillary and/or mandibular anterior teeth Pretreatment CBCT*: mm voxel size Direct measurement at surgery: Vernier caliper Dehiscence (d): > 2 mm from CEJ; fenestration (f): > 0 mm Compared direct measurements at surgery to pre-treatment CBCT measurements * KaVo 3D exam, Brugg, Switzerland * NewTom VG, NewTom, Verona, Italy Liangyan Sun et al. Accuracy of conebean computed tomography in detecting alveolar bone dehiscences and fenestrations. Am J Orthod Dentofac Orthop 2015;147: Liangyan Sun et al. Am J Orthod Dentofac Orthop 2015;147: New-Tom VG; 30 cm (12 inch) FOV; mm voxel; 5.4 seconds Measured bone height from CEJ; Mean difference: /- 4.1 mm 95 % limits of agreement Carestream CS 9300; 10 cm (4 inch) FOV; 0.18 mm voxel; 8 seconds Measured bone height from cusp tip; Mean difference: /- 1.4 mm 95% limits of agreement Table VI. Summary of direct and CBCT results for dehiscences and fenestrations by tooth type (N = 122) Tooth type Sites (n) Direct CBCT Dehis (n) Fenestr (n) Dehis (n) Fenestr (n) Maxilla Central incisor Lateral incisor Canine Mandible Central incisor Lateral incisor Canine Total Liangyan Sun et al. Accuracy of cone-bean computed tomography in detecting alveolar bone dehiscences and fenestrations. Am J Orthod Dentofac Orthop 2015;147: Comparisons: OHSU Studies Cadavers i-cat Age: Mean: 77 years Range: years Mix of posterior and anterior teeth (48): Maxilla: posterior: 4 (6%); anterior: 16 (25%) Mandible: posterior: 13 (20%); anterior: 32 (49%) CBCT considerations: O.3 mm voxel 0.5 mm slices Impact of embalming on scan resolution unknown relative to live tissues No movement artifacts Patients i-cat Age: Mean: 64 years Range: years Mix of posterior and anterior teeth (32): Maxillary: posterior: 2 (6%); anterior: 6 (18%) Mandible: posterior: 5 (15%); anterior: 19 (59%) CBCT considerations: 0.3 mm voxel 0.5 mm slices Presence of living cells: x-ray scatter from cells/organelles, high water content Patient movement during scanning Patients Carestream CS 9300 Age: Mean: 64 years Range: years Mix of posterior and anterior teeth (56): Maxillary: posterior: 28 (50%); anterior: 17 (30%) Mandible: posterior: 7 (12%); anterior: 4 (7%) CBCT considerations: 0.18 or 0.09 mm voxel 0.18 mm slices Presence of living cells: x-ray scatter from cells/organelles, high water content Patient movement during scanning Summary: Cadavers using Various CBCT Settings Accuracy: No difference in accuracy of measurements among scans having widely varying numbers of projection images (169 to 619) and differing arcs of rotation (360 o vs. 180 o ) Precision: Better with high resolution scans: Perhaps reserve for research purposes, or when high precision is desired Assessing alveolar bone height vs. thickness with CBCT: Bone height measurements correlate more closely to direct measurements than do bone thickness measurements 10

11 Summary: Live Patients and 0.3 mm Voxel Size With older CBCT unit (i-cat 17-19) and 0.3 mm voxel: Alveolar bone height was underestimated in CBCT images in areas where alveolar bone is thin Mean of 2.3 mm underestimation Particularly an issue with mandibular anterior teeth CBCT had notable limitations determining if thin bone or a bone defect is present 28% of teeth studied had either the presence of bone or a boney defect misidentified by one or both examiners Summary: Live Patients and 0.18 and 0.09 mm Voxel Size With newer CBCT unit and 0.18 or 0.09 mm voxel size scans: Alveolar bone height measurements were more accurate Mean of 0.5 mm underestimation No particular advantage of 0.09 vs mm voxel size CBCT still had notable limitations determining if thin bone or a bone defect is present 0.18 mm voxel size: 48% of teeth studied had bone present that was misidentified with CBCT as a defect by one or both assessors 0.09 mm voxel size: 26% of teeth had bone present that was misidentified with CBCT as a defect by one assessor or the other CBCT in Assessing Alveolar Bone Height: Are Looks Deceiving? Results of investigations at OHSU: Skulls/cadavers: Show best case scenario Measurements from CBCT had high accuracy and precision Caution needed translating results to clinical setting Clinical CBCT Imaging: Scans with small voxel size scans more accurate than with large voxel size Even with smallest voxel size, when imaging indicates a lack of alveolar bone along the facial surface of roots, one cannot know if thin bone is present, or if bone is absent Newer CBCT unit, smaller voxel size scans more accurate CBCT in Assessing Alveolar Bone Height: Are Looks Deceiving? Results of investigations at OHSU: Skulls/cadavers and bone height: Show best case scenario Measurements from CBCT had high accuracy and precision Caution needed translating results to clinical setting Clinical CBCT Imaging and bone height: Larger voxel size - less accuracy, precision Smaller voxel size, newer CBCT units Improved accuracy and precision, but will tend to underestimate Even with smallest voxel size, when imaging indicates a lack of alveolar bone, one cannot know if thin bone is present, or if bone is absent Reflections on CBCT Studies Clinical considerations: What is it that you want to view with CBCT imaging? For traditional radiographic records used in orthodontics: Moderate resolution likely adequate Large to medium field of view (medium if use digital lateral cephalogram) ALADA: larger voxel size ( mm), low dose scans Bone height: /- 5.5 mm (95% limit of agreement)* For assessing alveolar bone height: High resolution optimal Small field of view (single or both dental arches; localized area) ALADA: smaller voxel size (0.18 mm) Bone height : /-1.5 mm (95% limit of agreement)** * i-cat 17-19; 0.3 mm voxel scan ** Carestream CS 9300; 0.18 voxel scan Adam Timock (2010) Fort Collins, CO Lane Cook (2011) Garland, Tx Megan Miller (2013) Lancaster, OH Vanessa Browne (2014) Portland, OR Andrew Peterson (2017) Resident, Periodontology Mansen Wang Providence Health & Services Brion Benninger Acknowledgements Western Univ. of Health Sciences Jennifer Crowe Harjit Singh Sehgal Larry Doyle Terry McDonald OHSU Orthodontic Alumni Association 11